Intelligent typo warning method for blind typing of visually impaired people

Abstract

The present disclosure relates to an intelligent wrong-word warning method for blind typing by visually impaired persons, comprising: obtaining a first input word or symbol, and generating a reading audio signal of the first word or symbol; obtaining a second word or symbol within a predetermined field length input before and after the first word or symbol; combining the second word and the first word into search information in a predetermined order, matching the search information with a near-phonetic wrong-word rule library; when the search information and the matching result in the near-phonetic wrong-word rule library meet a first predetermined condition, generating wrong-word warning audio information; when the search information and the matching result in the near-phonetic wrong-word rule library meet a second predetermined condition, adding the first word to the output information. The above method can warn of wrong words during high-speed input by visually impaired persons, thereby improving input accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of barrier-free products, and in particular relates to an intelligent wrong-word warning method for visually impaired people to blindly type similar-sounding but different-words. Background Art

[0002] With the development of technology, more and more visually impaired people are using computers and other equipment to work. For example, some visually impaired people with professional skills can now participate in document typing, software programming and other tasks.

[0003] For these tasks, voice input can no longer meet the requirements of input accuracy, so keyboard typing, especially blind typing with pinyin, has become the inevitable choice for these visually impaired people. However, since visually impaired people cannot select characters or words with the same pinyin when using the input method, it is very easy to make mistakes. This brings instability to visually impaired people.

[0004] If the method of reading the words to be selected is adopted, it will take a long time to read the selected content, which brings great inconvenience to the input speed. Therefore, there is an urgent need in the prior art for a system that can take into account the success rate and accuracy of typing for visually impaired people and can provide early warning when visually impaired people make typing errors. Summary of the invention

[0005] The present disclosure is proposed based on the above-mentioned needs of the prior art. The technical problem to be solved by the present disclosure is to provide an intelligent typo warning system for blind typing by the visually impaired, so as to conveniently remind the visually impaired to improve the accuracy of input while maintaining efficient input.

[0006] In order to solve the above technical problems, the technical solutions provided by the present disclosure include:

[0007] An intelligent typo warning method for blind typing by visually impaired persons, characterized by comprising:

[0008] Acquire a first input character or symbol, and generate a reading audio signal of the first character or symbol; acquire a second character or symbol within a predetermined field length input before or after the first character or symbol; combine the second character and the first character into search information in a predetermined order, and match the search information with a near-phonetic wrong character rule library; when the search information and the near-phonetic wrong character rule library match the first predetermined condition, generate wrong character warning audio information; when the search information and the near-phonetic wrong character rule library match the second predetermined condition, add the first character to the output information.

[0009] Preferably, the first characters or symbols are characters or symbols determined by a pinyin input method according to an input command and a selection command.

[0010] Preferably, after obtaining the first input text or symbol, the first text or symbol is input into the first interface, and an audio signal for reading the first text or symbol is generated; obtaining the second text or symbol within a predetermined field length input before and after the first text or symbol includes obtaining the second text or symbol within a predetermined field length input before and after the first text or symbol in the first interface.

[0011] Preferably, the output information is used to input text or symbols into the second interface.

[0012] Preferably, the method further comprises: acquiring context information during input; selecting different near-phonetic typo rule libraries according to the context information; and different near-phonetic typo rule libraries correspond to different fields.

[0013] Preferably, when the search information and the matching result in the near-phonetic wrong character rule library meet the first predetermined condition, the correction information input by the user is received, and the first character or symbol is replaced by the correct character or symbol recorded in the near-phonetic wrong character rule library according to the user's correction information.

[0014] Preferably, obtaining the first input text or symbol includes obtaining the first input text or symbol by obtaining a keyboard input signal or a voice input signal.

[0015] Preferably, after replacing the first text or symbol with the correct text or symbol recorded in the near-phonetic wrong character rule library according to the user's correction information, recheck the other text or symbols before and after the replaced text or symbol according to the replaced content, and take the text or symbol within a predetermined range after the first text or symbol as the new first text or symbol; obtain new second text or symbol within a predetermined length before and after the new first text or symbol; combine the new second text and the new first text into new retrieval information in a predetermined order, and match the new retrieval information with the near-phonetic wrong character rule library; when the matching result between the new retrieval information and the near-phonetic wrong character rule library meets the first predetermined condition, generate new wrong character warning audio information; when the matching result between the new retrieval information and the near-phonetic wrong character rule library meets the second predetermined condition, add the new first text to the output information.

[0016] Preferably, an audio message of a wrong character warning is generated, the first text or symbol is used as the first field, the alarm field text is used as the second input field, and the first field and the second field are combined into an input signal of an audio generation module; and the audio generation module is used to generate an alarm audio message corresponding to the input signal.

[0017] Preferably, when the matching result between the search information and the near-phonetic wrong character rule library meets the first predetermined condition; receive the first correction information input by the user; generate an alternative list corresponding to the first correction information or symbol based on the user's correction information, and generate audio interpretation information corresponding to the alternative list one by one; receive the word selection information input by the user, and select the corresponding text or symbol in the alternative list according to the word selection information to replace the first text or symbol.

[0018] The present invention enables visually impaired people to understand the pronunciation of input text or symbols when inputting through voice reading, and perform preliminary screening based on the pronunciation. In addition, the input word can be judged whether it is a typo in terms of meaning through the context before and after the input word. In this way, high accuracy can be maintained when visually impaired people input text or symbols at high speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a flow chart of an intelligent wrong character warning method for blindly typing similar-sounding but different-sounding characters by a visually impaired person according to a specific implementation method of the present disclosure;

[0021] Figure 2 It is a schematic diagram of the link structure of the processing units through which data passes during the input process in the specific implementation mode of the present disclosure. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0023] In the description of the embodiments of the present disclosure, it should be noted that, unless otherwise clearly specified and limited, the term "connected" should be understood in a broad sense, for example, it can be a fixed connection, or a detachable connection, or an integral connection can be a mechanical connection, or an electrical connection can be a direct connection, or it can be indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present disclosure can be understood according to specific circumstances.

[0024] The terms "top", "bottom", "above", "lower", and "on" used throughout the description are relative to the relative positions of components of a device, such as the relative positions of the top and bottom substrates within a device. It is understood that devices are multifunctional regardless of their orientation in space.

[0025] To facilitate understanding of the embodiments of the present application, further explanation will be given below with reference to specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation on the embodiments of the present application.

[0026] Visually impaired people, also known as blind people, have many inconveniences in life. Especially when it comes to the widespread development of information technology and the emergence of a variety of visual media to support people's daily work, life, study and other basic needs of life, visually impaired people face many challenges in adapting to information technology.

[0027] In order to facilitate the use of high-tech products by people with visual impairments, barrier-free support applications such as screen reading software and voice input have emerged. However, on the computer or mobile phone input side, these applications still cannot fully cover all demand scenarios.

[0028] For example, when a visually impaired person needs to prepare reports, produce documents, etc. for various reasons, they need to have high input accuracy so that the works and products are highly readable and convey information more accurately.

[0029] The main difficulty faced by visually impaired people when inputting text or symbols is that their vision condition cannot support the screening and identification of input text by vision. Especially in the case of a large number of near-phonetic words and homophones in Chinese, it is very easy to input near-phonetic words and homophones incorrectly. If the method of selecting and explaining the meaning of near-phonetic words and homophones one by one is adopted, the input speed will often be greatly reduced. Therefore, there are obvious inconveniences for visually impaired people during the input process.

[0030] This specific implementation method relates to an intelligent typo warning method for blind typing by visually impaired persons, such as Figure 1-Figure 2 As shown, it is convenient for visually impaired people to find homophones and near-homophones during fast input and warn of typos.

[0031] The method for warning of wrong characters during blind typing provided in this specific implementation can be presented in an editor scenario or an input method scenario.

[0032] For example, in the editor application scenario, a text editor is provided for users to input text. Unlike ordinary editors, when an incorrect word that is homophonic with a correct word is input in a setting environment of a specified industry (such as the IT industry), a warning will be issued, such as a voice prompt "may be a typo". A typical word includes "homepage" being typed as "night watch" in the IT industry. Since the two have the same pronunciation, visually impaired people cannot determine whether the word is wrong if they only rely on screen reading. However, the method used in this specific implementation method can screen possible typos on homophones and near-phonetic words.

[0033] Another scenario includes the use of input methods. In this scenario, a text input method that can be installed by users can be set. Unlike ordinary input methods, when an incorrect word with the same pronunciation as the correct word is entered in a setting environment of a specified industry (such as the IT industry), a warning will be issued before the text is displayed on the screen, that is, when the text is completed in the text editing application, such as a voice prompt "may be a wrong word". Finally, the correct word confirmed by the user will be displayed on the screen.

[0034] For this reason, the intelligent wrong character early warning method for blind typing of the visually impaired comprises the following steps: Figure 1 Specifically, the intelligent typo warning method for blind typing by the visually impaired includes:

[0035] Step 1: Obtain a first input character or symbol, and generate a reading audio signal of the first character or symbol.

[0036] In this specific implementation, a preferred example can obtain the input information of the visually impaired person by setting an input interface. In addition, since the visually impaired person is normally invisible to the interface, the interface can also be presented only in the form of sound instead of display.

[0037] That is, when a visually impaired person inputs a character, the character information is stored in an intermediate unit, which temporarily stores the character information but does not send it to the application software. For example, when writing a document in an office software like Word, the intermediate unit temporarily stores the input character information but does not input the character information into the Word application, i.e., does not "go to the screen". This makes it easier to modify and select the appropriate correct character after the warning.

[0038] In this specific embodiment, the visually impaired can input text and symbols through a keyboard, including a physical keyboard, a keyboard formed by a software interface, or other similar input devices, or can input text or symbols through voice input. These input methods are actually unable to obtain the specific appearance of the text or symbol for the visually impaired, and therefore it is impossible to ultimately verify whether the input contents are accurate. In this step, after obtaining the above information, the audio of the input text or symbol is generated so that the visually impaired can hear the sounds and verify whether the pronunciation of the input text is accurate from the pronunciation. The method for generating audio can be implemented by calling a mature audio module, such as a voice module that comes with the operating system or a voice module provided by a third-party application.

[0039] However, it is not possible to determine whether the input text is correct or wrong just by reading the pronunciation, especially for homophones or near-phonetic words, which are particularly easy to cause confusion, and visually impaired people cannot judge the accuracy of these problems by vision. Therefore, in this case, storing these input contents through an intermediate unit is conducive to further verifying whether the input text and symbols are accurate.

[0040] The first character or symbol is a character or symbol determined by the pinyin input method according to an input command and a selection command.

[0041] Step 2: Obtain a second character or symbol within a predetermined field length input before or after the first character or symbol.

[0042] In this step, after a certain word or symbol is input, the correctness of the word or symbol is usually related to the context before and after it. Therefore, in addition to obtaining the information of the input word or symbol, it is also necessary to obtain the information of the related words or symbols within the predetermined range before and after it.

[0043] For example, in the IT field, when browsing the "home page" of a website, the most frequently used pages are "enter home page", "return to home page" and other related pages. When a user enters a word pronounced "shou ye", the related words before and after it are judged, and the information that assists in judging whether the word pronounced "shou ye" is accurate will be obtained.

[0044] Therefore, in this specific implementation, it is also necessary to obtain the second text or symbol within the predetermined field length input before and after the first text or symbol. According to different expressions, the length of the predetermined field length before and after can be set, and can be set to the length of the interval with the input word, or the number of the intercepted text before and after.

[0045] Or, the auxiliary judgment information can be used to judge whether the input text is accurate based on a fixed combination of common words in the context before and after the input text.

[0046] In this specific embodiment, further preferably, second characters or symbols within a predetermined field length input before and after the first character or symbol in the first interface are obtained. That is, only the characters and symbols within a predetermined length before and after the first character or symbol stored in the intermediate unit are obtained. Since the characters and symbols stored in the intermediate unit are characters or symbols that have not been entered on the screen, warning of possible errors in the characters or symbols during this process is beneficial to improving the accuracy of the characters entered on the screen after input screening.

[0047] Step 3: Combine the second character and the first character into a retrieval information in a predetermined order, and match the retrieval information with the near-sound misspelling rule library; when the matching result of the retrieval information and the near-sound misspelling rule library meets the first predetermined condition, generate misspelling warning audio information; when the matching result of the retrieval information and the near-sound misspelling rule library meets the second predetermined condition, add the first character to the output information.

[0048] In this step, the second character and the first character are combined into retrieval information in a predetermined order. The combination of the first character and the second character can directly combine the second character and the second character adjacent to each other, or the second character and the second character can be combined in other relationships, such as, but not limited to, the way of combining other characters between the first character and the second character, etc. By combining the information of the first character and the second character, common phrases can be combined, and it can be judged whether the input word belongs to a misspelled word or phrase according to the words before and after homophones and near-sound words.

[0049] For example, in the context of the IT field, the home page usually refers to the home page of a website. Therefore, common collocations include "home page browsing", "browse the home page", "enter the home page", "return to the home page", "top of the home page", "bottom of the home page", etc. The homophone of "home page" in Chinese is "stay up late", and the near-sound words are "teach", etc. These homophones and near-sound words cannot be collocated with the words such as "browse", "enter", "return", "top", "bottom", etc. before and after in the above context. Therefore, when the input vocabulary of "shouye" and the words before and after it are combined as the retrieval content, homophones and near-sound word misspellings can be found relatively quickly.

[0050] In this specific embodiment, the homophone or near-sound misspelling rule library can include, for example, commonly used homophones and near-sound words that are easily input incorrectly, as well as their fixed collocations. For example, the words usually fixedly collocated with "home page" generally include "browse", "enter", "return", "top", "bottom". The above phrases that are usually fixedly collocated with "home page" are entered in the homophone or near-sound misspelling rule library, so that when homophones and near-sound words included in the input are matched with the rule library, it can be checked whether the pre-entered characters belong to misspelled words, thereby warning of possible homophone and near-sound misspellings that may be entered.

[0051] The rule bases can be set to be multiple, for example, to match different industries respectively. Different industries correspond to different wording characteristics. Different rule bases are selected according to the industry to improve the accuracy of judgment.

[0052] In this specific implementation, the search content and the rule base are preferably matched by complete matching, that is, the search content and the fixed collocations entered in the rule base are completely corresponding, which can improve the speed and accuracy. However, according to actual needs, those skilled in the art can also use other matching methods besides complete matching, which is often determined according to the situation where typos are prone to occur in actual usage scenarios.

[0053] In this specific embodiment, whether the input content may be a typo is determined based on the correspondence between the search content and the rule base. For example, in one specific embodiment, when the search content matches the content included in the rule base, it is determined that the corresponding first text or symbol information may be a typo; and when the search content does not match the content included in the rule base, it is determined that the corresponding first text or matching information is not a typo. In another embodiment, it can also be set in a completely opposite manner. For example, when the search content matches the content included in the rule base, it is determined that the corresponding first text or symbol information is not a typo; and when the search content does not match the content included in the rule base, it is determined that the corresponding first text or matching information is a typo.

[0054] When it is identified that the first text or symbol information may be a wrong character, the first text or symbol and the wrong character prompt content are input into an audio generation module, such as an audio generation module provided by an operating system provider or an audio module provided by a third-party application provider, so as to generate an audio including the pronunciation of the first text or symbol and the wrong character information. For example, "Watching the night may be a wrong character".

[0055] Furthermore, in this case, a command from the user can be accepted, and the input text that may be a typo can be corrected or not corrected according to the user's command.

[0056] A preferred solution includes replacing the first word or symbol with the correct word or symbol recorded in the near-phonetic wrong word rule library according to the user's correction information. After replacing the first word or symbol with the correct word or symbol recorded in the near-phonetic wrong word rule library according to the user's correction information, recheck the other words or symbols before and after the replaced word or symbol according to the replaced content, and use the words or symbols within a predetermined range after the first word or symbol as the new first word or symbol; obtain the new second word or symbol within a predetermined length before and after the new first word or symbol; combine the new second word and the new first word into new search information in a predetermined order, and match the new search information with the near-phonetic wrong word rule library; when the matching result of the new search information and the near-phonetic wrong word rule library meets the first predetermined condition, generate new wrong word warning audio information; when the matching result of the new search information and the near-phonetic wrong word rule library meets the second predetermined condition, add the new first word to the output information.

[0057] Another preferred scheme includes generating an alternative list corresponding to the first correction information or symbol based on the user's correction information, and generating audio interpretation information corresponding to the alternative list one by one; the user inputs word selection information based on the audio interpretation information; receiving the word selection information input by the user, and selecting the corresponding text or symbol in the alternative list based on the word selection information to replace the first text or symbol.

[0058] When the matching structure meets the input condition, the text stored in the intermediate storage unit can be input into the application software according to the user's instruction, that is, displayed on the screen.

[0059] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only the specific implementation method of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. An intelligent typo warning method for blind typing by visually impaired persons, characterized in that: include: Acquire a first input character or symbol, and generate a reading audio signal of the first character or symbol; Obtaining a second character or symbol within a predetermined field length input before and after the first character or symbol; Combining the second character and the first character into search information in a predetermined order, matching the search information with a near-phonetic wrong character rule library; generating wrong character warning audio information when the search information and the near-phonetic wrong character rule library match a first predetermined condition; adding the first character to output information when the search information and the near-phonetic wrong character rule library match a second predetermined condition; When it is identified that the first character or symbol information may be a wrong character, the first character or symbol and the wrong character prompt content are input into the audio generation module; Receiving a command from a user, and correcting or not correcting the input text that may be a typo according to the command of the user; When the search information and the matching result in the near-phonetic wrong character rule library meet the first predetermined condition, receiving correction information input by the user, and replacing the first character or symbol with the correct character or symbol recorded in the near-phonetic wrong character rule library according to the user's correction information; Recheck other characters or symbols before and after the replaced character or symbol according to the replaced content, and use characters or symbols within a predetermined range after the first character or symbol as new first characters or symbols; Acquire a new second character or symbol within a predetermined length before and after the new first character or symbol; New search information is formed by combining the new second character and the new first character in a predetermined order, and the new search information is matched with a near-phonetic wrong character rule library; when the matching result between the new search information and the near-phonetic wrong character rule library meets a first predetermined condition, new wrong character warning audio information is generated; when the matching result between the new search information and the near-phonetic wrong character rule library meets a second predetermined condition, the new first character is added to the output information.

2. The intelligent typo warning method for blind typing by the visually impaired according to claim 1, characterized in that: The first characters or symbols are characters or symbols determined by the pinyin input method according to an input command and a selection command.

3. The intelligent wrong-word warning method for blind typing by the visually impaired according to claim 1, characterized in that: After acquiring the input first character or symbol, input the first character or symbol into the first interface, and generate a reading audio signal of the first character or symbol; Obtaining a second word or symbol within a predetermined field length input before or after the first word or symbol, includes obtaining a second word or symbol within a predetermined field length input before or after the first word or symbol in the first interface.

4. The intelligent wrong-word warning method for blind typing by the visually impaired according to claim 3, characterized in that: The output information is used to input text or symbols into the second interface.

5. The intelligent wrong-word warning method for blind typing by the visually impaired according to claim 3, characterized in that: The method further comprises: Get context information during input; Different near-phonetic mispronunciation rule bases are selected according to the context information; different near-phonetic mispronunciation rule bases correspond to different fields.

6. The intelligent wrong-word warning method for blind typing by the visually impaired according to claim 1, characterized in that: Acquiring the first input text or symbol includes acquiring the first input text or symbol by acquiring a keyboard input signal or a voice input signal.

7. The intelligent wrong-word warning method for blind typing by the visually impaired according to claim 1, characterized in that: Generate word warning audio information, use the first word or symbol as a first field, use the warning field word as a second input field, and combine the first field and the second field into an input signal of an audio generation module; An audio generation module is used to generate alarm audio information corresponding to the input signal.

8. The intelligent wrong-word warning method for blind typing by the visually impaired according to claim 1, characterized in that: When the matching result between the search information and the near-phonetic wrong character rule library meets the first predetermined condition; receiving first correction information input by a user; Generating an alternative list corresponding to the first correction information or symbol according to the user's correction information, and generating audio interpretation information corresponding to the alternative list one by one; Receive word selection information input by a user, and select corresponding words or symbols in the candidate list to replace the first words or symbols according to the word selection information.

Images